Method and device for analyzing pathology slide image

Abstract

The computing device according to an aspect comprises: at least one memory; and at least one processor, wherein the at least one processor is configured to: identify at least one tissue region from a pathology slide image by using an artificial intelligence model; generate spatial distribution-related information for the at least one tissue region on the basis of an image arithmetic operation of the at least one tissue region; and calculate an evaluation index associated with treatment responsiveness, on the basis of the spatial distribution-related information.

Description

Method and device for analyzing pathology slide images

[0001] The present disclosure relates to a method and device for analyzing a pathology slide image.

[0002] The field of digital pathology is a field that obtains histological information or predicts prognosis of a patient by using a whole slide image created by scanning a pathological slide image.

[0003] Recently, technologies have been developed to predict medical information by analyzing pathology slide images using artificial intelligence models. However, methods for analyzing pathology slide images with high accuracy and speed, as well as various biomarkers for predicting medical information, are still needed.

[0004] The present invention provides a method and device for analyzing pathology slide images. Furthermore, the present invention provides a computer-readable recording medium containing a program for executing the method on a computer. The technical problems to be solved are not limited to the technical problems described above, and other technical problems may exist.

[0005] A computing device according to one aspect comprises at least one memory; and at least one processor; wherein the at least one processor identifies at least one tissue region from a pathology slide image using an artificial intelligence model, generates spatial distribution-related information for the at least one tissue region based on an image operation for the at least one tissue region, and calculates an evaluation index associated with treatment responsiveness based on the spatial distribution-related information.

[0006] A method for analyzing a pathology slide image according to another aspect includes: identifying at least one tissue region from the pathology slide image using an artificial intelligence model; generating spatial distribution-related information for the at least one tissue region based on an image operation for the at least one tissue region; and calculating an evaluation index associated with treatment responsiveness based on the spatial distribution-related information.

[0007] Another aspect of a computer-readable recording medium includes a recording medium having recorded thereon a program for executing the above-described method on a computer.

[0008] Biomarkers derived from analysis of pathology slide images can be used to develop effective treatment strategies for tumors. Furthermore, they can help specialists prescribe more appropriate and effective personalized treatments to patients.

[0009] FIG. 1 is a diagram illustrating an example of a system for analyzing a pathology slide image according to one embodiment.

[0010] FIG. 2A is a configuration diagram illustrating an example of a user terminal according to one embodiment.

[0011] FIG. 2b is a configuration diagram illustrating an example of a server according to one embodiment.

[0012] FIG. 3 is a flowchart illustrating an example of a method for analyzing a pathology slide image according to one embodiment.

[0013] FIG. 4 is a diagram illustrating an example of identifying at least one tissue region from a pathology slide image using an artificial intelligence model according to one embodiment.

[0014] FIG. 5 is a diagram illustrating an example of the spatial distribution of an organizational area according to one embodiment.

[0015] FIG. 6 is a diagram illustrating an example of contact between a tumor region and a stroma region according to one embodiment.

[0016] Figure 7 is an example diagram of multiple sections according to one embodiment.

[0017] FIG. 8 is an example diagram of a pathology slide image showing multiple sections according to one embodiment.

[0018] FIGS. 9A to 9C are diagrams illustrating examples of calculating a fragmentation index according to one embodiment.

[0019] FIG. 10 is a diagram illustrating another example of a system for analyzing pathology slide images.

[0020] A computing device according to one aspect comprises at least one memory; and at least one processor; wherein the at least one processor identifies at least one tissue region from a pathology slide image using an artificial intelligence model, generates spatial distribution-related information for the at least one tissue region based on an image operation for the at least one tissue region, and calculates an evaluation index associated with treatment responsiveness based on the spatial distribution-related information.

[0021] The terms used in the examples are selected from widely used, current terms, as much as possible. However, these terms may vary depending on the intentions of those skilled in the art, precedents, the emergence of new technologies, etc. Furthermore, in certain cases, the applicant may arbitrarily select terms, in which case their meanings will be described in detail in the relevant description. Therefore, the terms used in the specification should be defined based on their intended meaning and the overall content of the specification, rather than simply their names.

[0022] When a part of the specification is said to "include" a component, this does not exclude other components, but rather implies the inclusion of other components, unless otherwise specifically stated. Furthermore, terms such as "unit" and "module" used throughout the specification refer to a unit that processes at least one function or operation, which may be implemented in hardware, software, or a combination of hardware and software.

[0023] Additionally, terms including ordinal numbers, such as "first" or "second," used in the specification may be used to describe various components, but the components should not be limited by the terms. The terms may be used to distinguish one component from another.

[0024] In one embodiment, a "pathology slide image" may refer to an image of a pathology slide that has been fixed and stained through a series of chemical treatments, such as for tissues removed from a human body. In addition, the pathology slide image may refer to a whole slide image (Whole Slide Image, WSI) that includes a high-resolution image of the entire slide, or may refer to a portion of the whole slide image, for example, one or more patches. For example, the pathology slide image may refer to a digital image captured or scanned by a scanning device (e.g., a digital scanner, etc.), and may include information about specific proteins, cells, tissues, and / or structures within the human body. In addition, the pathology slide image may include one or more patches, and histological information may be applied (e.g., tagged) to one or more patches through annotation.

[0025] "Medical information" may refer to any medically meaningful information that can be extracted from a medical image. For example, the medical information may include at least one of immune phenotype, genotype, biomarker score, tumor purity, RNA information, tumor microenvironment information, and treatment methods for cancer depicted in a pathology slide image.

[0026] Additionally, medical information may include, but is not limited to, the area, location, size of specific tissues (e.g., cancer tissue, cancer stromal tissue, etc.) and / or specific cells (e.g., tumor cells, lymphocytes, macrophages, endothelial cells, fibroblasts, etc.) within a medical image, diagnostic information of cancer, information related to the likelihood of a subject developing cancer, and / or medical conclusions related to cancer treatment.

[0027] Additionally, medical information may include not only quantifiable figures obtained from medical images, but also visualized figures, predicted figures, image information, and statistical information.

[0028] For example, medical information may be provided to a user terminal or output through a display device.

[0029] Below, embodiments are described in detail with reference to the attached drawings. However, the embodiments may be implemented in various different forms and are not limited to the examples described herein.

[0030] FIG. 1 is a diagram illustrating an example of a system for analyzing a pathology slide image according to one embodiment.

[0031] Referring to FIG. 1, a system for analyzing a pathology slide image (hereinafter, “system”) includes a user terminal (10) and a server (20). For example, the user terminal (10) and the server (20) may be connected via wired or wireless communication to transmit and receive various data between each other.

[0032] For convenience of explanation, FIG. 1 illustrates the system as including a user terminal (10) and a server (20), but the present invention is not limited thereto. For example, the system may include other external devices (not shown). Furthermore, the operations of the user terminal (10) and server (20), which will be described below, may be implemented by a single device (e.g., the user terminal (10) or the server (20)) or by multiple devices.

[0033] The user terminal (10) may be a computing device equipped with a display device and a device for receiving user input (e.g., a keyboard, a mouse, etc.), and including memory and a processor. Furthermore, the display device may be implemented as a touch screen and perform a function for receiving user input. For example, the user terminal (10) may be, but is not limited to, a notebook PC, a desktop PC, a laptop, a tablet computer, a smartphone, etc.

[0034] The server (20) may be a device that communicates with an external device (not shown) including a user terminal (10). For example, the server (20) may be a device that stores various data, including a pathology slide image, a bitmap image corresponding to the pathology slide image, information generated by analysis of the pathology slide image (e.g., information on at least one tissue region expressed in the pathology slide image, information on at least one biomarker expression, medical information related to the pathology slide image, etc.), and information on an artificial intelligence model used for analysis of the pathology slide image. Alternatively, the server (20) may be a computing device that includes a memory and a processor and has its own computing capability. When the server (20) is a computing device, the server (20) may perform at least some of the operations of the user terminal (10) that will be described later with reference to FIGS. 1 to 10. For example, the server (20) may be a cloud server, but is not limited thereto.

[0035] The user terminal (10) outputs pathology slide images and / or information generated by analysis of the pathology slides. For example, the user terminal (10) may output various information regarding at least one tissue region represented in the pathology slide image. Additionally, the user terminal (10) may output biomarker expression information.

[0036] A pathology slide image can refer to an image of a pathology slide that has undergone a series of chemical treatments, fixed, and stained for microscopic observation of tissues removed from the human body. As an example, a pathology slide image can refer to a whole slide image, which includes a high-resolution image of the entire slide. Alternatively, a pathology slide image can refer to a portion of such a high-resolution whole slide image.

[0037] Meanwhile, a pathology slide image may refer to an area divided into patches (or tiles) from the entire slide image. For example, a patch (or tile) may have a certain area size.

[0038] Additionally, a pathology slide image may refer to a digital image captured using a microscope, and may include information about cells, tissues, and / or structures within the human body.

[0039] Analysis of pathology slide images can identify biological elements (e.g., cancer cells, immune cells, tumor regions, etc.) depicted in the images. These biological elements can be used for histological diagnosis of disease, prediction of disease prognosis, and determination of treatment strategies.

[0040] The user terminal (10) can analyze a pathology slide image using an artificial intelligence model. For example, the user terminal (10) can individually detect (detection) numerous cells distributed on a pathology slide image and segment (segment) each region in the tissue using the artificial intelligence model. Furthermore, the user terminal (10) can analyze the detected cells and segmented regions using the artificial intelligence model and derive analysis results.

[0041] Hereinafter, with reference to FIGS. 2 to 10, an example of a user terminal (10) analyzing a pathology slide image will be described.

[0042] Meanwhile, for convenience of explanation, the user terminal (10) is described as performing all operations throughout the specification, but this is not limited to this. For example, at least some of the operations performed by the user terminal (10) may also be performed by the server (20).

[0043] FIG. 2A is a configuration diagram illustrating an example of a user terminal according to one embodiment.

[0044] Referring to FIG. 2A, the user terminal (100) includes a processor (110), a memory (120), an input / output interface (130), and a communication module (140). For convenience of explanation, only components related to the present invention are illustrated in FIG. 2A. Therefore, in addition to the components illustrated in FIG. 2A, other general-purpose components may be further included in the user terminal (100). In addition, it will be apparent to those skilled in the art that the processor (110), memory (120), input / output interface (130), and communication module (140) illustrated in FIG. 2A may be implemented as independent devices.

[0045] The processor (110) can process computer program commands by performing basic arithmetic, logic, and input / output operations. Here, the commands can be provided from memory (120) or an external device (e.g., a server (20), etc.). In addition, the processor (110) can generally control the operations of other components included in the user terminal (100).

[0046] The processor (110) can identify at least one tissue region from a pathology slide image using an artificial intelligence model. As an example, the tissue region may include a tumor region and a stroma region.

[0047] The processor (110) may generate spatial distribution-related information for at least one tissue region based on an image operation for at least one tissue region. At this time, the spatial distribution-related information may include at least one of area information of each of a plurality of sections included in at least one tissue region based on a boundary where a tumor region and a stroma region contact each other, area information of an area where a stroma region is extended by a preset length and an intersection area where the tumor region overlaps, or the number of fragments of a tumor region included in at least one tissue region.

[0048] As an example, the processor (110) may divide the tumor region and the stroma region into a plurality of sections based on a predetermined interval based on a boundary where the tumor region and the stroma region included in at least one tissue region contact each other, and may calculate the density of at least one target cell to be analyzed included in each of the plurality of sections. The processor (110) may set the analysis range in the direction of the stroma region and the direction of the tumor region based on the boundary. In addition, the processor (110) may generate a plurality of sections by dividing the analysis range based on the interval.

[0049] Specifically, the processor (110) can identify a first intersection area where the expanded area and the tumor area overlap based on an area where the substrate area is expanded by a first length and an area where the tumor area is contracted by a first length, and determine area information for an area corresponding to one of the plurality of sections based on the first intersection area.

[0050] In addition, the processor (110) may set an intersection area where the expanded area and the substrate area overlap based on an area where the tumor area is expanded by a first length and an area where the substrate area is contracted by the first length, and may determine area information for an area corresponding to one of a plurality of sections included in the substrate area based on the intersection area. In addition, the processor (110) may detect fragments of the tumor area.

[0051] Similarly, the processor (110) may identify a second intersection area where the area expanded by the second length and the area where the tumor area is contracted by the first length overlap, based on an area where the substrate area is expanded by a second length longer than the first length and an area where the tumor area is contracted by the second length, and may determine area information for an area corresponding to another section among a plurality of sections included in the tumor area based on the second intersection area.

[0052] The processor (110) may calculate an evaluation index associated with treatment responsiveness from at least one parameter calculated based on spatial distribution-related information. At this time, the evaluation index may include at least one of the density of the target cell in each of a plurality of sections generated based on the tumor-stromal border (TSB) (hereinafter, border or tumor-stromal contact border), the contact score (CTS) (hereinafter, tumor-stromal contact index), or the tumor fragmentation index (TFI) (hereinafter, fragmentation index). Each embodiment will be described below.

[0053] Thereafter, the processor (110) can calculate the density of at least one target cell included in each of the plurality of sections. For example, the processor (110) can calculate the density of tumor infiltrating lymphocytes (TILs) in each of the plurality of sections.

[0054] As an example, the processor (110) may detect one or more boundary grids including a boundary among a plurality of grids of a pathology slide image. In addition, the processor (110) may calculate the density of tumor-infiltrating lymphocytes in each of the plurality of sections included in the one or more boundary grids.

[0055] At this time, the processor (110) can predict the therapeutic responsiveness to the immuno-oncology agent of at least one tissue region based on the density of tumor-infiltrating lymphocytes. Specifically, the processor (110) can set a weight to the density of at least one target cell of analysis in at least one section among the plurality of sections.

[0056] As another example, the processor (110) may determine an intersection area where a substrate area extends by a preset length and a tumor area overlap, as described above, and calculate a tumor-stroma contact index based on a ratio of the area of ​​the intersection area to the area of ​​the tumor area.

[0057] As another example, the processor (110) can calculate a fragmentation index based on the ratio of the number of fragments in a tumor region to the area of ​​the tumor region.

[0058] Meanwhile, the processor (110) may output spatial distribution-related information of at least one tissue region identified from a pathology slide image. Furthermore, the processor (110) may output a report including spatial distribution-related information calculated based on the spatial distribution. Furthermore, the processor (110) may output a report including an evaluation index calculated based on the spatial distribution-related information. For example, the processor (110) may control a display device so that at least one of the aforementioned spatial distribution-related information, evaluation index, or report is output.

[0059] The processor (110) may be implemented as an array of a plurality of logic gates, or may be implemented as a combination of a general-purpose microprocessor and a memory storing a program that can be executed on the microprocessor. For example, the processor (110) may include a general-purpose processor, a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), a controller, a microcontroller, a state machine, etc. In some environments, the processor (110) may include an application-specific integrated circuit (ASIC), a programmable logic device (PLD), a field programmable gate array (FPGA), etc. For example, the processor (110) may refer to a combination of processing devices, such as a combination of a digital signal processor (DSP) and a microprocessor, a combination of a plurality of microprocessors, a combination of one or more microprocessors coupled with a digital signal processor (DSP) core, or any other such combination of configurations.

[0060] The memory (120) may include any non-transitory computer-readable recording medium. As an example, the memory (120) may include a non-permanent mass storage device such as a random access memory (RAM), a read only memory (ROM), a disk drive, a solid state drive (SSD), a flash memory, etc. As another example, the non-permanent mass storage device such as a ROM, an SSD, a flash memory, a disk drive, etc. may be a separate permanent storage device distinct from the memory. In addition, the memory (120) may store an operating system (OS) and at least one program code (e.g., a code for the processor (110) to perform an operation to be described later with reference to FIGS. 3 to 10).

[0061] These software components may be loaded from a computer-readable recording medium separate from the memory (120). This separate computer-readable recording medium may be a recording medium that can be directly connected to the user terminal (100), and may include, for example, a computer-readable recording medium such as a floppy drive, a disk, a tape, a DVD / CD-ROM drive, a memory card, etc. Alternatively, the software components may be loaded into the memory (120) through a communication module (140) other than a computer-readable recording medium. For example, at least one program may be loaded into the memory (120) based on a computer program (e.g., a computer program for the processor (110) to perform the operations described below with reference to FIGS. 3 to 10) that is installed by files provided by developers or a file distribution system that distributes installation files of applications through the communication module (140).

[0062] The input / output interface (130) may be a means for interfacing with a device (e.g., a keyboard, a mouse, etc.) for input or output that may be connected to or included in the user terminal (100). In FIG. 2A, the input / output interface (130) is illustrated as an element configured separately from the processor (110), but is not limited thereto, and the input / output interface (130) may be configured to be included in the processor (110).

[0063] The communication module (140) may provide a configuration or function for the server (200) and the user terminal (100) to communicate with each other via a network. In addition, the communication module (140) may provide a configuration or function for the user terminal (100) to communicate with other external devices. For example, control signals, commands, data, etc. provided under the control of the processor (110) may be transmitted to the server (200) and / or external devices via the communication module (140) and the network.

[0064] Meanwhile, although not illustrated in FIG. 2A, the user terminal (100) may further include a display device. Alternatively, the user terminal (100) may be connected to an independent display device via wired or wireless communication to transmit and receive data between the two devices. For example, pathology slide images, analysis information on pathology slide images, medical information, and additional information based on medical information may be provided to the user (30) via the display device.

[0065] FIG. 2b is a configuration diagram illustrating an example of a server according to one embodiment.

[0066] Referring to FIG. 2B, the server (200) includes a processor (210), a memory (220), and a communication module (230). For convenience of explanation, only components related to the present invention are illustrated in FIG. 2B. Therefore, in addition to the components illustrated in FIG. 2B, other general-purpose components may be further included in the server (200). Furthermore, it will be apparent to those skilled in the art that the processor (210), memory (220), and communication module (230) illustrated in FIG. 2B may be implemented as independent devices.

[0067] The processor (210) may obtain a pathology slide image from at least one of the internal memory (220), the user terminal (100), and another external device. The processor (210) may identify at least one tissue region from the pathology slide image using an artificial intelligence model, confirm the spatial distribution of at least one tissue region based on an image operation for at least one tissue region, calculate an evaluation index associated with treatment responsiveness from spatial distribution-related information calculated based on the spatial distribution, or transmit at least one of information on the identified tissue region, information on the spatial distribution of the tissue region, information on the spatial distribution-related information, and an evaluation index calculated from the spatial distribution-related information to the user terminal (100). In addition, the processor (210) may transmit at least one of the evaluation index or a report (e.g., including a predicted treatment responsiveness based on the evaluation index) to the user terminal (100).

[0068] In other words, at least one of the operations of the processor (110) described above with reference to FIG. 2A may be performed by the processor (210). In this case, the user terminal (100) may output information transmitted from the server (200) through a display device.

[0069] Meanwhile, since the implementation example of the processor (210) is the same as the implementation example of the processor (110) described above with reference to FIG. 2a, a detailed description thereof is omitted.

[0070] The memory (220) may store various data, such as pathology slide images and data generated according to the operation of the processor (210). In addition, the memory (220) may store an operating system (OS) and at least one program (e.g., a program required for the processor (210) to operate).

[0071] Meanwhile, the implementation example of the memory (220) is the same as the implementation example of the memory (120) described above with reference to FIG. 2a, so a detailed description is omitted.

[0072] The communication module (230) may provide a configuration or function for the server (200) and the user terminal (100) to communicate with each other via a network. In addition, the communication module (230) may provide a configuration or function for the server (200) to communicate with other external devices. For example, control signals, commands, data, etc. provided under the control of the processor (210) may be transmitted to the user terminal (100) and / or external devices via the communication module (230) and the network.

[0073] FIG. 3 is a flowchart illustrating an example of a method for analyzing a pathology slide image according to one embodiment.

[0074] The method illustrated in FIG. 3 is comprised of steps that are processed in time series in the user terminal (10, 100) or processor (110) illustrated in FIGS. 1 and 2A. Therefore, even if omitted below, the content described above regarding the user terminal (10, 100) or processor (110) illustrated in FIGS. 1 to 2A can also be applied to the method illustrated in FIG. 3.

[0075] Additionally, as described above with reference to FIGS. 1 to 2b, at least one of the steps of the method illustrated in FIG. 3 may be processed in the server (20, 200) or the processor (210).

[0076] At step 310, the processor uses the artificial intelligence model to identify at least one tissue region from the pathology slide image.

[0077] In the present disclosure, a pathology slide image may be a whole slide image or a portion of a whole slide image. Here, the portion may be referred to as a patch or a tile.

[0078] In one embodiment, a pathology slide image can be partitioned into multiple grids. As an example, the processor can partition the pathology slide image to generate multiple grids. The multiple grids can be formed in a grid shape. By generating multiple grids for the pathology slide image, the processor can efficiently organize the pathology slide image and easily manage and analyze data for each grid area.

[0079] In one embodiment, the processor can analyze a pathology slide image to identify a tumor region, a stroma region, a necrosis region, and a background region from the pathology slide image. Furthermore, the processor can classify a plurality of cells represented in the pathology slide image as at least one of tumor cells, lymphocytes, and other cells. The lymphocytes may include tumor-infiltrating lymphocytes. The processor can determine the area of ​​the identified tumor region, stroma region, necrosis region, and background region. Furthermore, the processor can determine the number of each of the plurality of cells.

[0080] In addition, biomarker expression information may include the number of positive (+) / negative (-) tumor cells for PD-L1 (programmed death-ligand 1), the number of positive (+) / negative (-) immune cells, the tumor proportion score (TPS), the combined positive score (CPS), etc. However, the biomarker expression information is not limited to what was described above, and an evaluation index associated with treatment responsiveness may include biomarker expression information.

[0081] FIG. 4 is a diagram illustrating an example of identifying at least one tissue region from a pathology slide image using an artificial intelligence model according to one embodiment.

[0082] Referring to FIG. 4, in one embodiment, the processor can identify at least one tissue region (431, 432) from a pathology slide image (430) using an artificial intelligence model (420). Specifically, the processor can input an image of a pathology slide (410) into the artificial intelligence model (420) to identify at least one tissue region (431, 432) included in the pathology slide image (430). The pathology slide image (430) may be an image of a pathology slide (410) that has been fixed and stained through a series of chemical treatments, such as tissue removed from a human body.

[0083] A tissue is a collection of specialized cells that perform a common function, and a specific area where these cells are clustered is called a tissue area. In one embodiment, at least one tissue area (431, 432) may include a tumor area (431) and a stroma area (432).

[0084] The tumor area (431) refers to an area where tumor cells exhibiting invasion are clustered, and the stromal area (432) refers to a peripheral area of ​​the tumor area (431) where tumor-related stromal changes, such as desmoplasia or aggregation of lymphatic cells, are observed.

[0085] The artificial intelligence model (420) may be a machine learning model. As an example, the artificial intelligence model may be trained to infer histological information for all or part of the pathology slide image (430) (e.g., at least one patch included in the pathology slide image (430). In this case, the histological information generated through the annotation task may be used to train the artificial intelligence model (420). For example, the histological information may include, but is not limited to, information about cells (e.g., tumor cells, lymphocytes, macrophages, dendritic cells, fibroblasts, endothelial cells, etc.) within the patch (e.g., the number of specific cells, information about the tissue in which specific cells are located). As another example, the artificial intelligence model (420) may be trained to infer characteristics of at least one of cells, tissues, or structures within the pathology slide image (430).

[0086] In one embodiment, the artificial intelligence model (420) can identify at least one tissue region (431, 432) by using histological information to determine information about the region in which cells are located in a pathology slide image (430).

[0087] Referring again to FIG. 3, at step 320, the processor may generate spatial distribution-related information for at least one tissue region based on image operations for at least one tissue region.

[0088] As an example, the processor can determine the spatial distribution of at least one tissue region based on an image operation on at least one tissue region.

[0089] The spatial distribution of at least one tissue region may include the shape and fragments of the at least one tissue region, the boundaries where different tissue regions contact each other, etc. In other words, the spatial distribution of the at least one tissue region means how the at least one tissue region is arranged in the pathology slide image and the positional relationship between different tissue regions.

[0090] In the present disclosure, the spatial distribution-related information may include information related to the spatial distribution of a tissue region acquired by the processor based on an image operation for at least one tissue region included in a pathology slide image. The spatial distribution-related information may include at least one of area information of each of a plurality of sections included in the at least one tissue region based on a boundary where a tumor region and a stroma region contact, area information of an area where the stroma region is extended by a preset length and an intersection area where the tumor regions overlap, the number of fragments of a tumor region included in the at least one tissue region, information on an analysis range set based on a boundary where a tumor region and a stroma region included in the at least one tissue region contact, information (e.g., location, area, etc.) on a boundary where a tumor region and a stroma region included in the at least one tissue region contact, location information of the tumor region, or location information of the stroma region.

[0091] Image operations may include at least one of, but are not limited to, image filtering, which processes an image by applying a specific kernel to the image; binarization, which divides the pixel values ​​of an image into two groups according to a threshold value; image transformation, which changes the size or shape of an image; image segmentation, which divides an image into multiple parts; and morphological operations, which manipulate the shape of an image by dilation, erosion, etc. The processor may perform operations on a pixel-by-pixel basis on a pathology slide image.

[0092] FIG. 5 is a diagram illustrating an example of the spatial distribution of an organizational area according to one embodiment.

[0093] In one embodiment, the processor can determine the spatial distribution of at least one tissue region (510, 520) based on image operations for at least one tissue region (510, 520).

[0094] As an example, the spatial distribution may include fragments of a tumor region (510) and / or a stroma region (520). That is, the processor may detect fragments of the tumor region (510) and / or the stroma region (520) based on an image operation for at least one tissue region (510, 520). For example, the processor may perform an image operation to analyze the identified tumor region (510) in preset units (e.g., pixels) to determine whether the tumor region (510) is composed of separate regions (i.e., fragments) or adjacent regions. That is, if the processor determines that the tumor regions (510) are separated by even one pixel, the tumor regions (510) may be detected as different fragments.

[0095] The processor can recognize the tumor region (510) as a fragment unit as a result of the image operation and generate spatial distribution-related information including the number of fragments in the tumor region. As illustrated in FIG. 5, the processor can detect four fragments in the tumor region (510) and generate spatial distribution-related information having a value of 4 as the number of fragments in the tumor region.

[0096] Meanwhile, when analyzing pathology slide images, only basic numerical values ​​such as the area of ​​the entire tumor region (510) and / or stroma region (520), the number of tumor cells, the number of immune cells, the density of tumor cells, and the density of immune cells are used as biomarkers, which has been raised as a problem in that it is limited in its use in predicting treatment responsiveness. However, according to the above-described embodiment of the present disclosure, an index reflecting the shape and degree of fragmentation of the tumor region (510) and / or stroma region (520) in the pathology slide image is proposed as a new biomarker, thereby improving the accuracy of predicting treatment responsiveness.

[0097] As another example, the spatial distribution may include a boundary (500) where a tumor region (510) and a stroma region (520) contact each other (hereinafter referred to as a “tumor-stroma contact boundary”). That is, the processor may detect information related to the tumor-stroma contact boundary (500) and / or the vicinity of the tumor-stroma contact boundary (500) based on image operations for at least one tissue region (510, 520).

[0098] FIG. 6 is a diagram illustrating an example of contact between a tumor region and a stroma region according to one embodiment.

[0099] In the present disclosure, the case of “contact between a tumor region (510) and a substrate region (520)” includes all cases where the outline of the tumor region (510) and the outline of the substrate region (520) are in contact, and cases where the tumor region (510) and the substrate region (520) do not overlap but are located within a preset adjacent distance.

[0100] Below, a process is described in which a processor performs image operations on at least one tissue region (610, 620) to detect information related to a tumor-stroma contact boundary.

[0101] FIG. 6 illustrates an area in which the substrate area (620) is extended by a preset length. In one embodiment, the processor may set an area in which the substrate area (620) is extended by a preset length. In addition, the processor may set an area (not shown) in which the tumor area (610) is contracted by a preset length. Specifically, the processor may set an expanded substrate area by expanding (expanding) the substrate area (620) by a preset length, and may set a contracted tumor area by contracting the tumor area (610) by a preset length. At this time, the processor may extract a first intersection area of ​​the expanded substrate area and the tumor area (610). As an example, the processor may set the substrate area (620) by a first length (e.g., 10 ) to set an expanded substrate region, and contract the tumor region (610) by a first length to set a contracted tumor region. Thereafter, the processor can extract an intersection of the difference between the expanded substrate region and the substrate region (620) and the difference between the tumor region (610) and the contracted tumor region, and determine this as the first intersection region.

[0102] In another embodiment, the processor may set a region (not shown) in which the tumor region (610) is expanded by a preset length. In addition, the processor may set a region (not shown) in which the substrate region (620) is contracted by a preset length. Specifically, the processor may set an expanded tumor region by expanding (expanding) the tumor region (610) by a preset length, and may set a contracted substrate region by contracting the substrate region (620) by a preset length. At this time, the processor may extract a first intersection region of the expanded tumor region and the substrate region (620). As an example, the processor may set the tumor region (610) by a first length (e.g., 10 ) to set an expanded tumor region, and contract the substrate region (620) by a first length to set a contracted substrate region. Thereafter, the processor can extract an intersection of the difference between the expanded tumor region and the tumor region (610) and the difference between the substrate region (620) and the contracted substrate region, and determine this as the first intersection region.

[0103] Meanwhile, in one embodiment, the processor may determine area information for a region corresponding to one of a plurality of segments included in the tumor region based on the first intersection region. In another embodiment, the processor may calculate an evaluation index associated with treatment responsiveness based on the first intersection region (e.g., the density of target cells in each of the plurality of segments generated based on the tumor-stroma boundary, a tumor-stroma contact index, etc.).

[0104] Although the example of the tumor region (610) and the substrate region (620) being expanded or contracted by a preset length was described above using FIG. 6, the processor may also extract the first intersection region and the second intersection region by expanding or contracting the tumor region (610) and the substrate region (620) by a preset width or a preset number of specific units (e.g., a preset number of pixels).

[0105] The regions of tissue adjacent to the tumor-stroma interface are crucial for tumor invasion and invasion, and are closely related to tumor progression and metastasis. Therefore, they must be accurately detected. According to one embodiment of the present disclosure, the regions of tissue adjacent to the tumor-stroma interface are detected based on image processing on pathology slide images, thereby improving accuracy.

[0106] Referring again to FIG. 3, at step 330, the processor calculates an evaluation index associated with treatment responsiveness based on the spatial distribution-related information.

[0107] Thereafter, the processor can calculate an evaluation index associated with treatment responsiveness based on spatial distribution-related information.

[0108] Figure 7 is an example diagram of multiple sections according to one embodiment.

[0109] In one embodiment, the processor can divide the tumor region and the stroma region into a plurality of sections (711 to 713, 721 to 723) based on preset intervals, based on the tumor-stroma contact boundary (701).

[0110] As an example, the preset interval is 10 can be. The preset interval can be the size of the largest cell, the size of the largest target cell, the average size of the target cells, or the average size of the cells (e.g., 7 8 inland ) can be set based on at least one piece of information, and the preset interval is 10 is not limited to and may be changed according to various embodiments. For example, the preset interval in the direction of the substrate region and the preset interval in the direction of the tumor region may be set to different values.

[0111] In one embodiment, the processor may set the analysis range (700) in the direction of the stromal region and the direction of the tumor region based on the tumor-stromal contact boundary (701). For example, the analysis range (700) may be 30 in the direction of the stromal region based on the tumor-stromal contact boundary (701). 30 in the direction of the tumor area based on the range and tumor-stroma contact boundary (701). It may mean a region that combines the ranges. However, the analysis range (700) is not limited to the aforementioned embodiment. As a non-limiting example, the analysis range (700) may include a tumor region and a stroma region at different distances from the tumor-stroma contact boundary (701).

[0112] In one embodiment, the first intersection region described above with reference to FIG. 6 may be any one of the plurality of sections (711 to 713, 721 to 723). That is, the first intersection region, in which the region where the substrate region is extended by the first length and the tumor region overlap, may be any one of the plurality of sections (711 to 713, 721 to 723) in the direction of the tumor region. Similarly, the first intersection region, in which the region where the tumor region is extended by the first length and the substrate region overlap, may be any one of the plurality of sections (711 to 713, 721 to 723) in the direction of the substrate region.

[0113] In another embodiment, the processor comprises a substrate region having a first length (e.g., 10 ) longer than the second length (e.g., 20 ) based on the region in which the tumor region is expanded by the second length and the region in which the tumor region is contracted by the first length, a second intersection region in which the region in which the tumor region is expanded by the second length and the region in which the tumor region is contracted by the first length overlap can be identified. Specifically, the processor can set the expanded region by expanding the substrate region by the second length, and can set the contracted tumor region by contracting the tumor region by the second length. Thereafter, the processor can extract an intersection of the difference between the substrate region expanded by the second length and the substrate region expanded by the first length, and the difference between the tumor region contracted by the first length and the tumor region contracted by the second length, and determine this as the second intersection region. The second intersection region determined in this way may be any one of a plurality of sections (711 to 713, 721 to 723) in the direction of the tumor region.

[0114] Likewise, the processor can identify a second intersection region in which the region expanded by the second length and the region contracted by the first length overlap, based on the region in which the tumor region is expanded by the second length and the region in which the substrate region is contracted by the second length, in order to determine which of the plurality of segments (711 to 713, 721 to 723) in the direction of the substrate region. Descriptions that overlap with the embodiment and principle of detecting a segment in the direction of the tumor region based on the second length will be omitted.

[0115] In one embodiment, the processor can calculate the density of at least one target cell included in each of the plurality of sections (711 to 713, 721 to 723). For example, the target cell may be a tumor-infiltrating lymphocyte. The processor can determine the area of ​​each of the plurality of sections (711 to 713, 721 to 723) included in the at least one tissue region based on an image operation for the at least one tissue region, based on a boundary where the tumor region and the stromal region contact each other. The processor can determine the number of target cells included in each of the plurality of sections (711 to 713, 721 to 723) among the detected target cells. Alternatively, the processor can determine the number of target cells included in each of the plurality of sections (711 to 713, 721 to 723) based on information (location, area) about the respective sections. The processor can determine the density of the cells to be analyzed for the section by using the area of ​​each section of the plurality of sections (711 to 713, 721 to 723) and the number of cells to be analyzed included in the section.

[0116] In one embodiment, the processor can detect one or more grids among a plurality of grids of a pathology slide image that include a tumor-stroma contact boundary (701). This is defined as a boundary grid.

[0117] Thereafter, the processor can calculate the density of immune cells in each of the plurality of sections (711 to 713, 721 to 723) included in one or more boundary grids. Furthermore, the processor can predict the therapeutic responsiveness of a patient from whom at least one tissue region has been sampled to an immunotherapy agent based on the density of immune cells. For example, the immune cells may be tumor-infiltrating lymphocytes.

[0118] Below, an embodiment is described in which a processor predicts therapeutic responsiveness to an immunotherapy agent based on the density of immune cells.

[0119] The processor can determine at least one of an immunophenotype or information associated with an immunophenotype of at least one region within the pathology slide image based on at least one of a density of immune cells in a plurality of sections (711 to 713) in the direction of a tumor region, or a density of immune cells in a plurality of sections (721 to 723) in the direction of a stroma region.

[0120] Specifically, the processor may determine the immune phenotype of at least some region in the pathology slide image as immune inflamed when the density of immune cells in the plurality of sections (711 to 713) within the tumor region is equal to or greater than a first threshold density. Furthermore, the processor may determine the immune phenotype of at least some region in the pathology slide image as immune excluded when the density of immune cells in the plurality of sections (711 to 713) within the tumor region is less than the first threshold density and the density of immune cells in the plurality of sections (721 to 723) within the stroma region is equal to or greater than a second threshold density. Furthermore, the processor may determine the immune phenotype of at least some region in the pathology slide image as immune desert when the density of immune cells in the plurality of sections (711 to 713) within the tumor region is less than the first threshold density and the density of immune cells in the plurality of sections (721 to 723) within the stroma region is less than the second threshold density.

[0121] At this time, the processor can determine an immunophenotype for each of a plurality of sections (711 to 713, 721 to 723) in the pathology slide image. Accordingly, if the most frequently included immunophenotype in the plurality of sections (711 to 713, 721 to 723) is immune activity, the processor can predict that the patient associated with the pathology slide image will respond to the immunotherapy (i.e., the patient is a responder). Conversely, if the most frequently included immunophenotype in the plurality of sections (711 to 713, 721 to 723) is immune exclusion or immune deficiency, the processor can predict that the patient associated with the pathology slide image will not respond to the immunotherapy (i.e., the patient is a non-responder).

[0122] In this way, by dividing the periphery of the tumor-stroma contact boundary (701) into multiple sections (711 to 713, 721 to 723) and analyzing each section, the understanding of tumor progression can be improved, and accordingly, an effective treatment strategy can be developed.

[0123] Meanwhile, 10 in the direction of the tumor area based on the tumor-stroma contact boundary (701) Inner section (711) and 20 Immune cells in the inner section (712) are highly correlated with treatment responsiveness (e.g., responsiveness to immunotherapy). In one embodiment, the processor can set a weight to the density of tumor-infiltrating lymphocytes in at least one of the plurality of sections (711 to 713, 721 to 723). As an example, 10 in the direction of the tumor region among the plurality of sections (711 to 713, 721 to 723) Inner section (711) and 20 A high weight can be set to the density of immune cells in the inner section (712). However, examples of setting weights in multiple sections (711 to 713, 721 to 723) are not limited thereto.

[0124] In one embodiment, the processor can predict treatment responsiveness by reflecting the weights.

[0125] FIG. 8 is an example diagram of a pathology slide image showing multiple sections according to one embodiment.

[0126] Referring to the pathology slide image (800) illustrated in FIG. 8, a first interface (810) indicating the currently displayed area in the pathology slide image (800), a second interface (820) capable of adjusting the resolution, and a third interface (830) describing the range of each of a plurality of sections are illustrated.

[0127] In one embodiment, the first interface (810) allows the user to select a portion of a pathology slide image to be reviewed. Furthermore, the second interface (820) allows the user to adjust the resolution of the pathology slide image by utilizing a zoom function. The third interface (830) allows the user to view information regarding multiple sections as described above. The Tumor-Stroma Boundary (TSB) depicted in the drawing may correspond to the tumor-stroma contact boundary.

[0128] In this way, the pathology slide image (800) can be analyzed more in depth as the processor generates an interface that provides the user with multiple sections of the tumor-stroma contact boundary and its surroundings.

[0129] In one embodiment, multiple sections in a pathology slide image (800) may be represented in different colors.

[0130] Referring again to FIG. 6, in one embodiment, the processor may determine an intersection region (630) where the substrate region (620) extends to a predetermined length and the tumor region (610) overlap. The overlapping of two regions means an area corresponding to the intersection of the two regions.

[0131] In one embodiment, the processor may calculate the tumor-stroma contact index based on a ratio of the area of ​​the intersection region (630) to the area of ​​the tumor region (610). For example, the processor may calculate the tumor-stroma contact index based on a ratio of the area of ​​the first intersection region where the tumor region (610) overlaps the area of ​​the tumor region (610) and the area of ​​the stromal region (620) extended by the first length.

[0132] A higher tumor-stroma contact index indicates a greater contact area between the tumor region (610) and the stroma region (620), which predicts a higher responsiveness to immunotherapy. Therefore, according to one embodiment of the present disclosure, a novel biomarker can be proposed that considers not only the area of ​​the tumor region (610) and the stroma region (620), but also the shape and distribution characteristics. Furthermore, medical information derived from this, such as the presence or absence of immunotherapy response, and the supporting evidence thereof, can be confirmed together.

[0133] FIGS. 9A to 9C are diagrams illustrating examples of calculating a fragmentation index according to one embodiment.

[0134] It is assumed that tumor regions with the same area are depicted in Figures 9a to 9c. At this time, it can be confirmed that the number of fragments in the tumor regions of Figures 9a to 9c are 1, 3, and 12, respectively.

[0135] In one embodiment, the processor may calculate a fragmentation index based on the ratio of the number of fragments to the area of ​​the tumor region. That is, the fragmentation index refers to the average number of fragments present per unit area.

[0136] Comparing Figures 9a to 9c, the tumor region in Figure 9a consists of a single mass and has the lowest fragmentation index. Furthermore, the tumor region in Figure 9b consists of three fragments and has a higher fragmentation index than the tumor region in Figure 9a. The tumor region in Figure 9c consists of 12 fragments and has the highest fragmentation index.

[0137] In one embodiment, the processor calculates the area of ​​each of the plurality of fragments constituting the tumor region and a preset reference area (e.g., 968.2 ) can be compared. The processor can determine the number of fragments in the tumor area by identifying the number of at least one fragment having an area greater than or equal to a preset reference area.

[0138] In one embodiment, the processor can determine whether each of a plurality of fragments constituting the tumor region is in contact with the stromal region. The processor can determine the number of fragments in the tumor region by identifying the number of at least one fragment in contact with the stromal region. For example, the processor can determine the number of fragments in the tumor region by identifying the number of at least one fragment that is in contact with the outline of the stromal region within the pathology slide image. The processor can determine the number of fragments in the tumor region by identifying the number of at least one fragment that is not overlapping with the stromal region but is located within a predetermined proximity distance within the pathology slide image. As an example, the processor can utilize the presence or absence of the first intersection region described above to identify the number of at least one fragment that is located within the predetermined proximity distance from the stromal region within the pathology slide image. That is, the processor can determine a tumor region that is not overlapping with the stromal region but has the first intersection region as a fragment located within the predetermined proximity distance.

[0139] The fragmentation index reflects the relative growth rates of the tumor and stroma regions, which can more accurately predict the therapeutic response to immunotherapy in cancer patients. As shown in Figure 9a, a tumor region composed of a single mass can be interpreted as having a faster growth rate than the tumor regions in Figures 9b and 9c.

[0140] As the processor calculates the fragmentation index, it is easy to determine the relative growth rate of the tumor region to the stromal region and the influence of the stromal region on the tumor region.

[0141] According to the aforementioned embodiments of the present disclosure, predicting treatment responsiveness to immunotherapy can help specialists prescribe more appropriate and effective personalized treatments to patients. Furthermore, this can ultimately contribute to improving patient survival rates.

[0142] FIG. 10 is a diagram illustrating another example of a system for analyzing pathology slide images.

[0143] Referring to FIG. 10, the system (1000) is an example of a system and network for preparing, processing, and reviewing slide images of tissue samples using an artificial intelligence model.

[0144] According to various embodiments of the present disclosure, the method described above with reference to FIGS. 2A to 10 may be performed by at least one or a combination of a user terminal (1022, 1023), an image management system (1030), an AI-based biomarker analysis system (1040), a laboratory information management system (1050), and a hospital or laboratory server (1060).

[0145] The scanner (1021) can acquire a digitized image from a tissue sample slide generated using a tissue sample of a subject (1011). For example, the scanner (1021), the user terminal (1022, 1023), the image management system (1030), the AI-based biomarker analysis system (1040), the laboratory information management system (1050), and / or the hospital or laboratory server (1060) can each be connected to a network (1070) such as the Internet via one or more computers, servers, and / or mobile devices, or can communicate with the user (1012) via one or more computers and / or mobile devices.

[0146] The user terminals (1022, 1023), the image management system (1030), the AI-based biomarker analysis system (1040), the laboratory information management system (1050), and / or the hospital or laboratory server (1060) may generate, or otherwise acquire from another device, one or more tissue samples, tissue sample slides (pathology slides), digitized images of tissue sample slides (pathology slides), or any combination thereof of the subject (1011). In addition, the user terminals (1022, 1023), the image management system (1030), the AI-based biomarker analysis system (1040), the laboratory information management system (1050), and / or the hospital or laboratory server (1060) may acquire any combination of subject-specific information, such as the age, medical history, cancer treatment history, family history, past biopsy records, or disease information of the subject (1011).

[0147] A scanner (1021), a user terminal (1022, 1023), an AI-based biomarker analysis system (1040), a laboratory information management system (1050), and / or a hospital or laboratory server (1060) may transmit digitized pathology slide images, subject-specific information, and / or analysis results of digitized pathology slide images to an image management system (1030) via a network (1070). The image management system (1030) may include a storage for storing received images and a storage for storing analysis results.

[0148] In addition, according to various embodiments of the present disclosure, an artificial intelligence model learned and trained to predict at least one of information about at least one cell, information about at least one region, information related to a biomarker, medical diagnosis information, and / or medical treatment information from a pathology slide image of a subject (1011) may be stored and operated in a user terminal (1022, 1023), an image management system (1030), an AI-based biomarker analysis system (1040), etc.

[0149] As described above, a computer device according to one embodiment of the present disclosure can extract the spatial distribution of tissue regions from pathology slide images and automatically calculate an evaluation index from the spatial distribution. Therefore, this can replace the analysis method in which a pathologist directly examines only a portion of the tissue distribution across the entire pathology slide image. Furthermore, the prediction accuracy and analysis speed of a patient's treatment response to immunotherapy can be improved.

[0150] Meanwhile, the above-described method can be written as a program that can be executed on a computer, and can be implemented on a general-purpose digital computer that runs the program using a computer-readable recording medium. In addition, the structure of the data used in the above-described method can be recorded on a computer-readable recording medium through various means. The computer-readable recording medium includes storage media such as magnetic storage media (e.g., ROM, RAM, USB, floppy disk, hard disk, etc.) and optical reading media (e.g., CD-ROM, DVD, etc.).

[0151] Those skilled in the art will appreciate that the present invention can be implemented in modified forms without departing from the essential characteristics of the above-described invention. Therefore, the disclosed methods should be considered illustrative rather than restrictive. The scope of the claims, not the foregoing description, is defined by the scope of the patent, and should be interpreted to encompass all differences within the scope equivalent thereto.

Claims

1. At least one memory; and comprising at least one processor; At least one processor, Identifying at least one tissue region from a pathology slide image using an artificial intelligence model, Generating spatial distribution-related information for at least one tissue region based on an image operation for at least one tissue region, A computing device that calculates an evaluation index associated with treatment responsiveness based on the above spatial distribution-related information.

2. In paragraph 1, A method in which the above spatial distribution-related information includes at least one of area information of each of a plurality of sections included in the at least one tissue area based on a boundary where a tumor area and a stroma area contact each other, area information of an area in which the stroma area is extended by a preset length and an intersection area in which the tumor area overlaps, or the number of fragments of the tumor area included in the at least one tissue area.

3. In paragraph 1, At least one processor, Based on the boundary where the tumor area and the stromal area included in at least one tissue area contact each other, the tumor area and the stromal area are divided into a plurality of sections based on a preset interval, A computing device that calculates the density of at least one target cell of analysis included in each of the plurality of sections.

4. In paragraph 3, At least one processor, Detecting one or more boundary grids including the boundary among a plurality of grids of the above pathology slide image, A computer device that calculates the density of tumor-infiltrating lymphocytes in each of the plurality of sections included in the one or more boundary grids.

5. In paragraph 4, At least one processor, A computer device that sets a weight to the density of at least one target cell of analysis in at least one section among the plurality of sections.

6. In paragraph 3, At least one processor, Based on the region in which the substrate region is expanded by the first length and the region in which the tumor region is contracted by the first length, a first intersection region in which the expanded region and the tumor region overlap is identified, A computer device that determines area information for an area corresponding to one of a plurality of sections included in the tumor area based on the first cross-section area.

7. In paragraph 6, At least one processor, Based on the region in which the substrate region is extended by a second length longer than the first length and the region in which the tumor region is contracted by the second length, a second intersection region is identified in which the region extended by the second length and the region in which the tumor region is contracted by the first length overlap, A computer device that determines area information for an area corresponding to another section among a plurality of sections included in the tumor area based on the second cross section.

8. In paragraph 3, At least one processor, Based on the region in which the tumor region is expanded by the first length and the region in which the substrate region is contracted by the first length, an intersection region in which the expanded region and the substrate region overlap is set, A computer device that determines area information for an area corresponding to one of a plurality of sections included in the substrate area based on the above cross-section area.

9. In paragraph 1, At least one processor, Determine the intersection area where the substrate area is extended by a preset length and the tumor area overlaps, A computer device that calculates a tumor-stroma contact index based on the ratio of the area of ​​the intersection area to the area of ​​the tumor area.

10. In paragraph 1, At least one processor, Detecting fragments of a tumor area contained in at least one tissue area, A computer device that calculates a fragmentation index based on the ratio of the number of fragments to the area of ​​the tumor region.

11. A step of identifying at least one tissue region from a pathology slide image using an artificial intelligence model; A step of generating spatial distribution related information for at least one tissue region based on an image operation for at least one tissue region; and A method for analyzing a pathology slide image, comprising: a step of calculating an evaluation index associated with treatment responsiveness based on the spatial distribution-related information.

12. In paragraph 11, A method in which the above spatial distribution-related information includes at least one of area information of each of a plurality of sections included in the at least one tissue area based on a boundary where a tumor area and a stroma area contact each other, area information of an area in which the stroma area is extended by a preset length and an intersection area in which the tumor area overlaps, or the number of fragments of the tumor area included in the at least one tissue area.

13. In paragraph 11, The above calculating steps are: A step of dividing the tumor region and the stroma region into a plurality of sections based on a preset interval based on a boundary where the tumor region and the stroma region included in the at least one tissue region contact each other; and A method comprising: calculating the density of at least one target cell of analysis included in each of the plurality of sections.

14. In paragraph 13, The steps for calculating the above evaluation index are: A step of detecting one or more boundary grids including the boundary among a plurality of grids of the pathology slide image; and A method comprising: calculating the density of tumor-infiltrating lymphocytes in each of the plurality of sections included in the one or more boundary grids.

15. In paragraph 14, The step of calculating the density of the above tumor-infiltrating lymphocytes is: A method comprising: setting a weight to the density of at least one target cell of analysis in at least one section among the plurality of sections.

16. In paragraph 13, The above generating steps are: A step of identifying an intersection area where the expanded area and the tumor area overlap based on an area where the substrate area is expanded by a first length and an area where the tumor area is contracted by the first length; and A method comprising: a step of determining area information for an area corresponding to one of a plurality of areas included in the tumor area based on the above cross-section area; 17. In paragraph 13, The above generating steps are: A step of setting an intersection area where the expanded area and the substrate area overlap based on an area where the tumor area is expanded by a first length and an area where the substrate area is contracted by the first length; and A method comprising: a step of determining area information for an area corresponding to one of a plurality of sections included in the substrate area based on the above cross-section area; 18. In paragraph 11, The above calculating steps are: A step of determining an intersection area where the substrate area is extended by a preset length and the tumor area overlaps; and A method comprising: calculating a tumor-stroma contact index based on a ratio of the area of ​​the intersection area to the area of ​​the tumor area.

19. In paragraph 11, The above generating steps are: A step of detecting a fragment of a tumor region included in at least one tissue region; The above calculating steps are: A method comprising: calculating a fragmentation index based on the ratio of the number of fragments to the area of ​​the tumor region.

20. A computer-readable recording medium recording a program for executing the method of Article 11 on a computer.

Images